Lost Bitcoin Collider Pool

[WanderingPhilospher]

Anyone interested in a "Lost" Bitcoin Collider Pool?

A pool where people could throw some hashing power at it when they aren't using their CPU/GPU to try and find what is considered lost bitcoin.

The pool would not be for trying to find every wallet/address with a balance, just the coins that are deemed lost.

So question one, what would you say would be the requirements/consideration of "lost" bitcoin?

Question two, would anyone have the know how of how to put a pool together, programming wise?

I would host the server and pay for the startup of it, and the upkeep.  I just don't have that kind of brain muscle...I am a tinkerer, shade tree, but I
know nothing of servers and making a program communicate.

I guess the pool would be like a mining pool, people get a percent of any lost bitcoin based on ranges searched or some hash rate. But either way, there
would have to be checks/verification to ensure people aren't sending fake hash/ranges searched.

Think of it as a better upgrade, with different twists and methods compared to the LBC. I see interest in the LBC (what got me to think of this today), but it
seems "dead" or barely limping along.  From what I've read, the program your CPU/GPU uses to search with isn't the best. 
Ideally, the new pool would have different search programs that people could choose to use; think Bitcrack (mods), VanitySearch (mods), archaic python scripts, etc.

I have ideas on how it could work but first want to see if there is any interest AND if anyone would have an inkling on how to set up the "mining" pool. If so,
we could brainstorm as it progresses.  If not, no biggie.

Ideas, suggestions, comments?

[1715662494]

1715662494

[1715662494]

1715662494

[1715662494]

1715662494

[WanderingPhilospher]

Reserved for OP...

[WanderingPhilospher]

Reserved for OP...

[NotATether]

I think the mining pool's work server can be done in a scripting language like Python, for coordinating the work from the clients. For obvious reasons and client key searching program will have to be written in C or C++.

Now I mentioned a scripting language, because maybe it's possible to save costs by hosting it on some cloud's serverless function like AWS Lambda. I just got my AWS account but I don't have enough time to work on a side project like this.

[o_e_l_e_o]

So question one, what would you say would be the requirements/consideration of "lost" bitcoin?

I don't think there is any reliable way to identify any bitcoin as "lost".

We not infrequently see transactions spending coins which haven't moved in over 10 years, which many people would have considered lost. We even saw an early bitcoin miner sign a message last year from a hundred addresses all holding bitcoin which has been untouched since 2009/2010. Those bitcoin still haven't moved, but the signed message proves that they aren't lost. Since inactivity is not a valid metric, there is no way to prove bitcoin have been lost. Even if someone says that they owned an address and lost the keys, there is no way to verify that. The only way to verify ownership is to sign a message, which would mean the bitcoin are not actually lost, and there are plenty of reasons someone would lie about losing their keys, for example in an unfortunate boating accident.

[WanderingPhilospher]

So question one, what would you say would be the requirements/consideration of "lost" bitcoin?

I don't think there is any reliable way to identify any bitcoin as "lost".

We not infrequently see transactions spending coins which haven't moved in over 10 years, which many people would have considered lost. We even saw an early bitcoin miner sign a message last year from a hundred addresses all holding bitcoin which has been untouched since 2009/2010. Those bitcoin still haven't moved, but the signed message proves that they aren't lost. Since inactivity is not a valid metric, there is no way to prove bitcoin have been lost. Even if someone says that they owned an address and lost the keys, there is no way to verify that. The only way to verify ownership is to sign a message, which would mean the bitcoin are not actually lost, and there are plenty of reasons someone would lie about losing their keys, for example in an unfortunate boating accident.

So you don't think their is no way to "verify" coins are lost?  I agree with what you said. Even after x amount of years without any movement, maybe the owner is hodling or still trying to figure out their priv key/password/etc.

Many talk about the first coins mined by Satoshi "lost" unless he comes back around and still has the wallet.dat or private key info.  Others say that if an owner can't recover password info for their .dat file and sell it, and it's on the open market, they are lost.  And some say only the burner addresses are considered "lost" because they can not be recovered because some of the burner addresses are made up and have no checksum and therefore cannot be recovered. Lost versus unrecoverable seem to be two different things to me.

If you had to come up with set criteria for deeming coins "lost", what would it be? 

[WanderingPhilospher]

I think the mining pool's work server can be done in a scripting language like Python, for coordinating the work from the clients. For obvious reasons and client key searching program will have to be written in C or C++.

Now I mentioned a scripting language, because maybe it's possible to save costs by hosting it on some cloud's serverless function like AWS Lambda. I just got my AWS account but I don't have enough time to work on a side project like this.

Python eh?  Do you mean for the server AND assigning work, or just assigning work? I can fumble around decently in python. Do you have a link to an example?

Etar built a complete package in PureBasic which is really good.  But, it limits the key searchable to just one at a time.  Maybe I should spend the money on a PureBasic license and dig into the code more and then see if their is a way to merge it with an existing search program like bitcrak or vanity or any other popular tools.

For the #64 pool, that is what is being done.  You download and run one program and that program contains the server stuff and the actual program to search keys with. You can use bitcrack or vanity. It also has verification built in via generating a random key in the search range and you must report it back or the program doesn't consider that range searched.

[NotATether]

Python eh?  Do you mean for the server AND assigning work, or just assigning work? I can fumble around decently in python. Do you have a link to an example?

Both of them, it's blessed with really robust urllib3 and requests libraries, and I'd say most of the code would be related to networking. I don't have any examples unfortunately, but you'll need some backup storage for the work on file/database/or etcd key-value store  thing in case the server crashes so you can just continue where you left off.

Etar built a complete package in PureBasic which is really good.  But, it limits the key searchable to just one at a time.  Maybe I should spend the money on a PureBasic license and dig into the code more and then see if their is a way to merge it with an existing search program like bitcrak or vanity or any other popular tools.

I don't think he ever released the source for it (gmaxwell PM'ed me a long time ago to try to convince him to show me at least, but that never worked out) so I think the PureBasic license is a waste of money.

[o_e_l_e_o]

Many talk about the first coins mined by Satoshi "lost" unless he comes back around and still has the wallet.dat or private key info.

Presumed lost, sure. Provably lost, nope.

Others say that if an owner can't recover password info for their .dat file and sell it, and it's on the open market, they are lost.

Again, there is no proof these coins are lost. They could have set a password they know would be impossible to brute force and are scamming users by selling the wallet, which they will sweep at a later date. Or perhaps the wallet doesn't belong to them in the first place.

And some say only the burner addresses are considered "lost" because they can not be recovered because some of the burner addresses are made up and have no checksum and therefore cannot be recovered.

This is probably the closest we can come to coins which are lost for the time being. The chances of anyone currently knowing the private key to burner addresses such as 1BitcoinEaterAddressDontSendf59kuE is so astronomically small as to essentially be zero. However, these coins are still not provably lost, as there exists a chance, however small, that someone could stumble upon or brute force the private key to these addresses as some point in the future.

If you had to come up with set criteria for deeming coins "lost", what would it be?

The only coins which are provably lost are ones which cannot be spent. Such an example would be coins sent to OP_RETURN outputs. But if they cannot be spent, then that defeats the purpose of what you are trying to do.